You are on page 1of 22

Journal of Molecular Catalysis A: Chemical 138 Ž1999.


Hydroformylation of ethene with triphenylphosphine modified rhodium catalyst: kinetic and mechanistic studies
G. Kiss

, E.J. Mozeleski b, K.C. Nadler c , E. VanDriessche c , C. DeRoover


Corporate Research Laboratories, Exxon Research and Engineering, Annandale, NJ 08801, USA b Exxon Chemical, Annandale, NJ 08801, USA c Exxon Chemical International, B-1831 Machelen, Hermeslaan 2, Belgium Received 9 March 1998; accepted 29 April 1998

Abstract The kinetics of ethene hydroformylation with triphenylphosphine modified rhodium catalyst has been investigated. The reaction is first order in rhodium. The kinetic order for ethene is also one for most of the reaction conditions investigated, however, the combination of low phosphine and high olefin concentration leads to a saturation in ethene. The kinetic effects of CO and triphenylphosphine are also a function of their concentrations: both inhibit the reaction above a threshold value, while below the threshold a positive kinetic response can be observed. Hydrogen concentration has little effect on the rate of hydroformylation. The olefin hydrogenation selectivity is generally low and can be described as a linear function of the H 2rCO partial pressure ratio. Hydrogenation of the product propionaldehyde cannot be detected by GC. Based on a refined Wilkinson mechanism wC.A. Tolman, J.W. Faller, Mechanistic studies of catalytic reactions using spectroscopic and kinetic techniques, in: L.H. Pignolet ŽEd.., Homogeneous Catalysis with Metal Phosphine Complexes, Plenum Press, New York, 1983.x, a kinetic model has been developed which fits all experimental observations and can be used to predict reaction rates and selectivities over a wide range of reaction conditions. q 1999 Elsevier Science B.V. All rights reserved.
Keywords: Hydroformylation; Oxo reaction; Phosphine modified rhodium catalyst; Ethene; Kinetics; Mechanism

1. Introduction Worldwide hydroformylation Ž or oxo. capacity in 1993 exceeded 6 million tonsryear w2x. Approximately 60% of that capacity is based on the low pressure oxo process Ž LPO. using phosphine-modified rhodium carbonyls as catalysts. LPO is used to convert light olefins Ž mainly C 2 and C 3 ., CO, and H 2 to C 3 and C 4 aldehydes Ž Eq. Ž 1.. . Oxo aldehydes are mainly converted to C nG 8 alcohols and used in the manufacture of polyvinyl chloride Ž PVC. plasticizers and detergents. R–CH5CH 2 q CO q H 2 ™ R–CH 2 –CH 2 –CHO q R–CH Ž CHO . –CH 3 R s H or CH 3 Ž1. The most widely used ligand in LPO is triphenylphosphine Ž PPh 3 . . The pioneering work of Evans et al. w3,4x, Yagupsky et al. w5x and Brown and Wilkinson w6x has established that ŽI. in the

Corresponding author. Fax: q1-908-730-3198; e-mail:

1381-1169r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 1 - 1 1 6 9 Ž 9 8 . 0 0 1 6 6 - 6


G. Kiss et al.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176

Scheme 1. The main steps of the Heck–Breslow hydroformylation mechanism.

RhrPPh 3-catalyzed hydroformylation the stable HRhŽ CO.Ž PPh 3 . 3 monocarbonyl hydride complex is a direct catalyst precursor, ŽII. the major Rh-component present under reaction conditions is the dicarbonyl hydride HRhŽCO. 2 Ž PPh 3 . 2 , and Ž III. the main steps in the catalytic cycle follow the Heck–Breslow mechanism w7–10x developed for the Co-catalyzed oxo reaction Ž Scheme 1. . In order to explain the effect of PPh 3 concentration on product selectivities, Wilkinson proposed two different paths: a dissociative one and an associative one. The former is initiated by the dissociation of CO from the dicarbonyl hydride, and the catalytic cycle follows the 16r18-electron path. The latter is initiated by the association of the olefin with the hydride leading directly to an 18-electron alkyl intermediate via a 20-electron transition state. Later spectroscopic studies w11–25x further refined the details of the mechanism and today it is widely accepted w1,2x that the likely kinetic path is the one which involves 16- and 18-electron species only, i.e., the dissociative mechanism. While the main features of the mechanism have been well established by identifying spectroscopically observable intermediates, the kinetics of the process have not been thoroughly studied w2x. Although there are some kinetic studies w26–32x which address the effect of several process variables with the goal of developing a kinetic model, only one of them w26x offers an analysis correlating propylene hydroformylation kinetic data with a reaction mechanism. Other publications either offer empirical models w27–32x or focus on a limited number of variables and address only certain aspects Žmostly phosphine effect. of the reaction w33–43x. This paper reports the results of a parametric study of the hydroformylation of ethene with triphenylphosphine-modified Rh catalyst. A kinetic model has been developed by correlating the kinetic data with a refined Wilkinson mechanism w1x.

2. Experimental 2.1. Materials Solvents Ž anhydrous toluene, tetraetheneglycol dimethyl ether, dioctyl phthalate. , methylcyclohexane as GC internal standard, all from Aldrich; 2,2,4-trimethyl-1,3-pentanediol monoisobutyrate Ž or Texanol. from Eastman or ACROS, and isoamyl butyrate from ACROS were thoroughly deaerated but otherwise were used as received. All gases Ž 1:1 H 2rCO, methane, and ethene from MG Industries or Air Liquide., triphenylphosphine Ž from Aldrich. , and RhŽ acac.Ž CO. 2 Ž from Strem or Aldrich. were used as received. All organic compounds were checked for purity by GC Ž Perkin Elmer Autosystem or Hewlett-Packard 5890, Hewlett-Packard Pona column. . For the kinetic measurements Rh stock solutions were prepared by using the appropriate solvent. The stock solution was made by weighing Ž 0.1 mg accuracy. the solvent, RhŽ acac.Ž CO. 2 Ž acac s acetylacetonato., and PPh 3 ŽPrRh ratio of ca. 10. into a bottle. A vigorous evolution of CO suggested the formation of the RhŽ acac.Ž CO.Ž PPh 3 . complex. Catalyst solutions for the hydroformylation of ethene with the desired composition were prepared by weighing the Rh stock solution, additional phosphine, and solvent.

through a high precision pressure-regulator valve. Mass balances were established for all organic components and were typically 100 " 5%. . Blank experiments were carried out to test the efficiency of the injection method. PPh 3 . The configuration of all units were identical. equipped with speed-controlled mechanical stirrers. Aliquot samples of the product were prepared under nitrogen for GC analysis by adding known amounts of methyl cyclohexane as a GC internal standard. The reaction was initiated by injecting ethene into the reactor and bringing the pressure up to 1000 kPa with syngas. Kiss et al. Rate and selectivity values obtained in two different reactors at the same condition were statistically indistinguishable.Ž CO. The reaction was stopped by rapid cooling of the reactor typically after 80–95% olefin conversion. The reactors were equipped as in the batch units.3. Reproducibility was checked by carrying out at least two runs at each condition.G. Mass transfer rates for CO. stock solution. . The Rh charge at each condition was adjusted to ensure that the liquid phase concentration of the reactants would be above the 95% level of the gas–liquid equilibrium values. and the solvent. After cooling the reaction mixture. except for PPh 3 for which the mass balances were typically 100 " 20%. kinetic experiments were carried out in 500 ml autoclaves Ž Autoclave Engineers. at the reaction temperature. Approximately 15 mmol of ethene was also charged into a 75 ml injection bomb which then was mounted into the syngas feed-line of the reactor. The constant unit pressure was maintained with a back-pressure regulator.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 157 2. All volumes in the system were measured by gas volumetric methods.2. In these blank experiments the temperature and pressure in the PVT tank and the reactor equilibrated within 2 min or less. The pressure in the reactor was maintained constant by making up the consumed syngas from the PVT tank. H 2 . The reactors and feed vessels were equipped with certified Omega digital thermocouples and pressure gauges. The absence of mass transfer limits was verified by variable stirrer-speed experiments. The solution was charged into the reactor under nitrogen. 65 ml of catalyst solution was prepared in a vacuum atmospheres glove-box from the RhŽ acac.ŽPPh 3 . All units were equipped with on-line gas chromatographs Ž MTI or Hewlett-Packard. Batch kinetic experiments In a typical experiment. Apparatus The batch kinetic experiments were performed in 300 or 500 ml autoclaves Ž Autoclave Engineers. to ca. NMR and IR experiments were carried out under identical conditions to verify the formation of the equilibrium mixture of rhodium hydrides with a general formula of HRhŽCO. 500 kPa. The conversion was followed by recording the pressure drop in the PVT tank. a gas sample was taken into a 50 ml stainless steel bomb. The effluent flow was measured with wet-test meters ŽAmerican meter or Ritter. The reactors were connected to their gas feed vessels Ž PVT tanks. The reactor then was flushed and pressurized with syngas Ž a 1:1 mixture of H 2 and CO. x ŽPPh 3 . The catalyst was preformed in syngas Ž approximately 500 kPa. which sampled both the feed and the effluent. The temperature of the reactors was maintained within "18C. 2. The continuous-flow Ž continuous stirred tank reactor or CSTR. 4yx Ž x s 1 and 2. The liquid product was removed under nitrogen. carried out in the 300 ml autoclave.. The gas feed components were fed individually via Brooks mass-flow controllers which were calibrated in-house. and C 2 H 4 in tetraglyme at different stirrer speeds were also measured using literature methods w44x. The conversions and product selectivities were determined from the mass balance for ethene based on GC analysis of the final liquid and .

q1 ½ polefin Ž t . can be obtained by using the initial turnover frequency if the initial partial pressure and the Henry’s constant of ethene are known. k. Ž 2. and Ž2. Another way of analyzing the raw kinetic data is based on the analytical solution of the differential mass balance for the olefin Ž integral method. From Eq. R s universal gas constant. time data by approximating the differential Žd n olefinrd t . Ž7. t s constant Ž D p PVTrDt . the reaction rate can be expressed as follows: yd n olefinrd t s k Rh olefin Vliq . The rate of reaction was calculated by assuming that the gas consumption. t with the finite difference Ž D n olefinrDt .r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 gas phase. Kiss et al. polefin Ž t s 0 . From Eq. In the text this data analysis will be referred to as ‘differential method’. and the substrate Ž olefin. The turnover frequency Ž TOF. r Ž total PVT pressure drop . The rate of olefin conversion was calculated from the PVT pressure vs. t and utilizing the fact that for the consumed olefin D n olefin s constant D p PVT : Ž d n olefinrd t . . where Vgas s free gas volume in the reactor. 5 s yk Rh t Ž8. For a reaction which is first order in both the catalyst ŽRh.158 G. Ž5. reaction time plot.s liquid volume. and thus the pressure drop in the PVT tank. Ž4. and k s reaction rate constant. Then from Eqs. is a linear function of conversion. and gas volume.s kn Rh olefin where V liq. where polefin and Holefin are the partial pressure and the Henry’s constant of the olefin. Ž6. other annotations as in Eqs. t ( Ž D n olefinrDt . For a reaction which is first order in Rh and olefin: y ž d n olefin dt / ž sy Vgas RT q V liq .: F ln where Fs Vgas Holefin RTVliq .: TOF s k olefin s kpolefinrHolefin Ž5. the rate constant. Ž 6. t i i i Ž3. The later was determined at each reaction temperature from blank experiments. . This assumption is valid as long as the pressure of the reactor is constant. Ž7. – Ž 6. respectively. The initial pressure of the PVT tank was obtained by extrapolating to time zero on a pressure vs.. used in the kinetic analysis was defined as the rate of olefin conversion by 1 mol of rhodium and expressed as moles ethene converted to aldehydes per mole of Rh per second: TOF s y ½ž d n olefin dt /5 1 n Rh Ž2. . T s temperature. liquid product weight. where constants Ž total mole of aldehyde product. Holefin /ž d polefin dt / s k Rh polefin Holefin Vliq .

and CH 4 Ž used as GC internal standard and inert diluent. they allowed independent calibration for each gas component. The only other product detected is ethane.7%. 2. The initial partial pressure of ethene in the reactor was between approximately 80 and 160 kPa. The experiments were carried out at 1008C. The catalyst solution was prepared and charged into the reactor as described earlier for the batch experiments. Ž9. at low PPh 3 concentrations Ž. time plots ŽFig. since the TOF is independent of wRhx. and 100 " 5% for ethene. Results and discussions 3. the data are reproducible.4 vs. Calibrations were based on certified master blends prepared by Scott Specialty Gases or Air Liquide. Kiss et al. approximately 99. Unexpectedly.4% selectivity. 4 s y Ž k Rh rF . . WCOT Ultimetal with FID detection. ln Ä 1 y conversion Ž t . Also.2–0.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 159 Since polefin Ž t . are linear with a zero intercept. 3. t s constant t The lnÄ 1 y conversionŽ t . time plot therefore is linear if the reaction is first order in olefin. The Rh concentration was measured by atomic absorption spectroscopy.1. suggesting first order kinetics for ethene.4. s 1 y conversion Ž t . vs. H 2 . the first order kinetic regime . the reaction is first order in Rh. At the end of the experiment the composition of the catalyst solution and the degree of oxidation of the phosphine ligand were determined by 31 P NMR and by GC Ž HP5890. The GC analyses were performed by MTI’s M200 micro GC or a custom modified HP 6890.. The hydrogenation turnover frequency was determined from the rate of formation of ethane.3. Ž 10. and PPh 3 was first tested in batch reactors. polefin Ž t s 0 . The rate constant can be calculated from its slope if the free gas and liquid volumes. The hydroformylation selectivity at these conditions is high. The hydroformylation turnover frequency was calculated as an average from the mass balances for ethene and propanal. and 1000 kPa total pressure with a H 2rCO ratio of one.. 2. Section 2. Since both chromatographs sampled at constant Žatmospheric. CSTR kinetic experiments The continuous-flow hydroformylation of ethene was carried out in 500 ml Autoclave Engineers’ zipperclaves. 1. as described in Section 2.G. The lnŽ1 y conversion. 50 mmolrl. including the product aldehyde. Mass balances for each component were obtained independently of each other and typically were 100 " 10% for CO and H 2 . 50 mmolrl the turnover frequency vs. At PPh 3 concentrations higher than approx. CO. C 2 H 4 . Parametric study of reaction Õariables The kinetic response to solvents and to changes in the concentration of ethene. Interestingly. were individually fed via Brooks mass-flow controllers. pressure. As shown in Fig. a saturation kinetics can be observed for ethene. and the partial pressure and Henry’s constant for ethene at the reaction temperature are known. The gaseous feed components. ethene conversion plots ŽFig. 3. are also linear for more than three half lives which is also consistent with a first order in ethene Žcf.3. forming at 0.

160 G. The effect of PPh 3 will be discussed later in more detail. This saturation kinetics in olefin hydroformylation is not without precedent. Integral method plot for the hydroformylation of ethene. Although mass-transfer limitation can lead to similar kinetic response w45x. it can be excluded in our experiments because: Ž I. Using constant partial pressure of CO. 2. the ethene concentration at which the kinetic saturation occurs becomes lower with lower PPh 3 concentration.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Fig. Kiss et al. and the reaction can be accelerated by increasing the H 2rCO ratio in the reactor. for 1-decene by Divekar et al. Hydroformylation rate of ethene as a function of conversion at three different Rh concentrations. and Ž II. can be expanded to higher ethene concentrations. saturation. and for 1-dodecene by Bhanage et Fig. At the observed maximum gas-uptake rate the calculated gas concentration for all reactants is within 5% of the gas–liquid equilibrium level. or even substrate inhibition kinetics with RhrPPh 3 catalyst has also been reported for 1-hexene by Deshpande and Chaudhari w31x. . 1. The TOF is independent of the concentration of Rh. Thus. w29x.

The fractional kinetic order for propylene found by the Montedison group w26x may be the result of an operation in the intermediate kinetic regime. Ethene saturation at low PPh 3 concentrations suggests a rate-limiting step before which there is a quasi-equilibrium involving the olefin. Data obtained in the present study are also consistent with a first order Rh dependence Ž Fig. 3 . al. .G. Saturation kinetics in ethene hydroformylation at low PPh 3 concentrations. while with excess ligand it is one for a variety of substrates and ligands w15. but with no excess PPh 3 present. Similar kinetic response has been described by Brown and Wilkinson w6x for the hydroformylation of 1-hexene in benzene at 258C with HRhŽCO.ŽPPh 3 . In the dissociative 16r18-electron mechanism proposed by Fig. 4. In their study the maximum TOF does depend on the concentration of Rh. . w27x. Hydroformylation rate of ethene as a function of Rh concentration at constant PPh 3 concentration. 4. 3. Kiss et al.26–31x.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 161 Fig. the kinetic order for Rh is less than one. at their conditions Žno excess ligand. However.

make this latter scenario less likely. As illustrated by Fig. show that at the same reaction Fig.. since it has been found that the dissociation of CO or phosphine is much faster than the rate of catalytic hydroformylation. and with it reaction rate and product selectivity. 5. The data plotted in Fig. x ŽPPh 3 .162 G. 3yx Ž x s 1–2.e.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Tolman and Faller w1x the earliest such equilibrium step is the formation of HRhŽC 2 H 4 . while at low olefin concentrations the limiting rate is olefin addition to HRhŽ CO. 1. a comparison between Figs. 2 ŽPPh 3 . 2 Ž L. At constant partial pressure of CO the determining factor thus becomes the concentration of the phosphine ligand.17. . respectively. or CO dissociation from HRhŽCO. the reaction rate is increased at low conversion Ž zero order regime.. 3yx Ž x s 1–3. intermediates as well. Furthermore. The effect of H 2 rCO ratio in ethene hydroformylation at low PPh 3 concentrations. The build-up of a non-acyl intermediate has been reported from in situ CIR-FTIR studies of reacting phosphine-modified Rh catalyst systems w20x. Our results are in full agreement with this prediction. 3 and 5 clearly shows that the first order regime expands with increasing phosphine concentration.. 5. depends on the relative concentrations of CO and PPh 3. the turnover frequency is little affected by changes in the H 2rCO ratio at high conversion. x Ž PPh 3 . 4yx x Ž x s 1–4. x Ž PPh 3 . By increasing the rate of hydrogen activation the rate limiting step shifts back to that earlier stage of the catalytic cycle. 2 Ž L s phosphine ligand. Our data support Tolman’s proposal regarding a shift in the rate-limiting step.46x it should be clear that the distribution of Rh intermediates at different COrphosphine ligand ratios.Ž CO. 3.21x. however. In contrary. Kiss et al. NMR kinetic data w16. but can also involve the alkyl wC 2 H 5 –RhŽ CO. using the same catalyst composition. 2 . for example. in the ethene first order regime. They also proposed w20x that the shift from first to zero order in olefin is due to a shift in the rate-limiting step from CO insertion in the alkyl intermediate to CO dissociation from HRhŽ CO. The authors have assigned the observed IR-spectrum to an 18-electron alkyl complex. Based on the dissociative mechanism of the reaction w1. which latter complex is the dominating species under typical oxo conditions with RhrPPh 3. The changing kinetic response suggests that the rate of hydrogen activation is comparable to the rate of an earlier step in the mechanism. Tolman and Faller have proposed w1x that at the saturation limit the rate limiting step is H 2 activation. i. The corresponding rate-limiting steps are olefin and CO insertion or H 2 activation. x Ž PPh 3 . by increasing the H 2rCO ratio ŽFig. 4yx x or acyl wC 2 H 5 CO– RhŽCO.

The second order rate constant is a monotonous function of the phosphine concentration. on the other hand.32–34x. would show a monotonous inhibition function for the phosphine. At our reaction conditions.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 163 condition the TOFrwethenex is the same as long as the phosphine concentration is the same. A maximum in the reaction rate vs.49. and iso-butene w32x.0 wt. The positive kinetic response to the phosphine at low phosphine concentrations has been proposed w32x to be due to the break-up of Rh dimers thus providing a higher concentration of the catalytically active mononuclear rhodium hydrides. w26x reported a similar value. Although it is customary in the hydroformylation literature to report kinetic and selectivity data as a function of the PrRh ratio w2. phosphine concentration function has also been reported for the hydroformylation of propylene w47x. Catalytic activity as a function of PPh 3 concentration in the first order regime of ethene hydroformylation with RhrPPh 3 catalyst. obtained at 80 and 1008C. and the kinetic behavior reflects the trapping of a 16-electron intermediate by the phosphine. 270 and 480 kPa of CO. the catalytic activity is inhibited by PPh 3 Ž Fig. 6. is present. . At high phosphine concentrations the dominating species are mononuclear w1x. too. In the hydroformylation of propylene d’Oro et al. . We have found rather that the TOFrwethenex values are independent of the PrRh ratio. 7.G.41 and y0. as was mentioned earlier. the reaction rate has a maximum at approximately 30 mmolrl phosphine concentration Ž Fig. the kinetic order for the phosphine is between y0. only the first kinetic order regime can be observed for ethene if the phosphine concentration is higher than 0. w26x. as long as other reaction variables are constant. In the ethene zero order regime Ž low ethene conversion. . The mechanistic implications will be discussed in more detail later in Section 3.2. 1-hexene w38–40x. The effect of PPh 3 concentration can be best characterized by defining two regimes depending on the kinetic order of ethene. 6.65. y0. It is therefore quite possible that at low olefin concentrations these olefins. and from 2.5 to 15. and solely depend on the concentration of the phosphine. From our data. we have not found any correlation between those values when excess phosphine ŽPrRh) 3.05 molrl. Kiss et al. Fig.% PPh 3 . It can explain the large shift in the kinetic order of PPh 3 in the low phosphine concentration regime. In the ethene first order regime Ž high conversion data. It has to be mentioned that these literature data represent initial rates and thus were measured at high olefin concentrations. This observation is also in agreement with the conclusion of an earlier report by d’Oro et al. .

7. and 1-dodecene w27x with HRhŽ CO. The kinetic effect of CO and H 2 was investigated in continuous-flow reactors as described in Section 2. the kinetic order for CO is positive in a narrow range of low CO partial pressures but it becomes negative after going through a maximum reaction rate. Maximum observed rate as a function of PPh 3 concentration in ethene hydroformylation with RhrPPh 3 catalyst. 2 Ž PPh 3 .r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Fig. while the rate sensitivity toward pCO changes in the opposite direction. while the partial pressure of either CO or H 2 was changed using methane as an inert balance to maintain constant pressure. 1-octene w28x. 8. It is known w1. The effect of CO on the hydroformylation rate of ethene at different PPh 3 concentrations.4. the maximum is shifted toward higher pCO values at higher PPh 3 concentrations. As shown in Fig. 1-decene w29x. It should be noticed on Fig. Kiss et al.18x that under hydroformylation conditions the dominating Rh-hydride component is HRhŽ CO.18x via the Fig. Namely. that the position of the maximum on the CO partial pressure axis. In these experiments all other variables were kept constant.17.Ž PPh 3 . and the rate response to the same D pCO is a function of the PPh 3 concentration. . Similar kinetic response to CO has been reported for the hydroformylation of 1-hexene w39x. 8.3–6. 2 . and both CO and PPh 3 undergo rapid exchange w17.164 G. 8. 3 catalyst.

2 and HRhŽ CO. 3 at low catalyst concentration. 8. or bind an ethene molecule. 9.. 10. reported first order for the hydroformylation of 1-hexene with HRhŽ CO. 8 will contribute to any significant degree. These two coordinatively unsaturated species are believed to be the entry points of the two prevalent kinetic paths in the phosphine-modified Rh hydroformylation of olefins Ždepicted as Path A and B in Fig. .G. shows a first order dependence on the partial pressure of H 2 ŽFig. in the ethene first order regime. is a function of the PPh 3rCO ratio. respectively. and thus their contribution to olefin conversion. . the hydroformylation rate is not affected by changing hydrogen partial pressure Ž Fig. The effect of H 2 partial pressure on the hydroformylation rate of ethene with RhrPPh 3 catalyst. .2. Thus. 2 Ž PPh 3 .1 At high PPh 3 concentrations. i. They can be trapped by CO or PPh 3 and driven back to the corresponding coordinatively saturated hydrides. Kiss et al. however. corresponding 16-electron intermediates.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 165 Fig. and a positive but decreasing order at higher catalyst concentration. w1x. A similarly changing kinetic order was also reported for 1-decene w29x and allyl alcohol w30x. . The rate of ethene hydrogenation at the same conditions. Literature data regarding the kinetic order of H 2 in the hydroformylation of olefins may seem contradicting. Invariably.ŽPPh 3 . The overall rate is a linear combination of the rates of the two parallel paths. and the other two paths are negligible under the conditions of our investigation. initiating a catalytic cycle. concluded that the kinetic order for H 2 is essentially zero w26x. x ŽPPh 3 . on the other hand. The relative concentration of the two 16-electron intermediates. Brown and Wilkinson w6x. Chapter 3.Ž PPh 3 . The rate-limiting step in the hydroformylation of ethene therefore must be prior to H 2 activation. 9. HRhŽ CO. a detailed study of propylene hydroformylation with RhrPPh 3 catalyst ŽPrRh) 3.e. based on published equilibrium data w1x and our kinetic analysis Žcf. 3yx Ž x s 0–3. we propose that only the routes shown in Fig. in all studies the kinetic order of H 2 is higher than zero when the PPh 3 concentration is low and tends to zero when the 1 Note: Although there are four possible conversion paths starting with the corresponding 16-electron complexes with a general formula of HRhŽCO. It will be shown later that the observed kinetic response to CO can be reconciled with the above mechanism. . while first order was published for 1-hexene w31x and 1-dodecene w27x with the same catalyst.

Temperature effect in the hydroformylation of ethene with RhrPPh 3 catalyst. are significantly different from that of measured in tetraglyme. .r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Fig.56:1. Literature reports suggest increased rates when the solvent is changed from non-polar Fig. on the other hand.166 G. The effect of H 2 partial pressure on the hydrogenation rate of ethene with RhrPPh 3 catalyst.% PPh 3 are nearly the same. 5 wt. respectively. 11. phosphine concentration is high. the kinetic orders for both ethene and H 2 are dependent on the concentration of PPh 3 Ž cf. and discussion on the effect of ethene. dioctyl phthalate. Texanol. tetraglyme. . Kiss et al.2. and have a ratio of 0.% PPh 3 . 5 wt. 1 MPa COrH 2 Ž1:1. . covering a wide range of PPh 3 concentrations. 10. and isoamyl butyrate at 1008C. 3 and 5. Figs. 1 MPa COrH 2 Ž1:1..73:0. In our experiments. The mechanistic implications will be discussed in Section 3. The hydroformylation rates of ethene in toluene. The reaction rates measured in n-butanal and propanal at 808C.

. Their data were obtained from empirical fits of the experimental data. Substantial solvent effects have also been reported in the hydroformylation of allyl alcohol w35x and 1-octene w33x. In our study we have not found spectroscopic evidence for solvent coordination to the metal. Thus.48x and 22 kcalrmol for 1-butene w16x are relatively close to our values.1 " 0. and 1-dodecene. Kiss et al.6 and 19 kcalrmol reported for propylene w26. Temperature effect in the hydrogenation of ethene with RhrPPh 3 catalyst. 12. 11. The 20. and hydrogenation Ž Fig. The range of operating conditions covered in these continuous-flow runs is given in Table 1. The activation energy for ethene hydroformylation obtained from batch experiments at 460 kPa CO and 180–360 mmolrl PPh 3 concentration is somewhat higher.8 to 28 kcalrmol for the hydroformylation of 1-hexene. Divekar et al. 12. 1-decene. published activation energies ranging from 11. of ethene was tested in tetraglyme between 80 and 1108C at 210 kPa partial pressure of CO and 93 mmolrl concentration of PPh 3.9 kcalrmol. The later two papers suggest metal–solvent interactions.G. We did not observe similar effect in the batch hydroformylation of ethene or propylene in toluene. w27x.4. The effect of temperature on the rate of hydroformylation Ž Fig. as well. but increasing ethane selectivity at higher temperatures indicates higher activation energy for hydrogenation than for hydroformylation in the batch results.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 167 Fig. The kinetic data for both reactions were obtained from the same CSTR experiments. 19. Bhanage et al. hydrocarbons to oxygenates. in the hydroformylation of propylene the product butanal has been reported w42x to increase the reaction rate in benzene. 3.2 kcalrmol.2 kcalrmol. thus the activation energies can be model dependent. than for hydrogenation Ž 21. however.. Kinetic model for the hydroformylation and hydrogenation of ethene The database for the kinetic model development was generated in CSTR experiments as described in Section 2.2. which may explain their wide variance and some of the unusually low values. The activation energy for hydroformylation has been found to be lower Ž 18. Literature values for the activation energy in the hydroformylation of olefins scatter considerably. The kinetic data are summarized in the Appendix A. w29x and Deshpande and Chaudhari w31x.

2. i. the initiation of the catalytic cycle by the dissociation of CO or PPh 3 from HRhŽCO.058 Maximum 110. x Ž PPh 3 .16x that the concentration of binuclear Rh complexŽ es. x Ž PPh 3 .. x Ž PPh 3 . since it is reasonable to assume that energetically the associative route would be strongly disfavored over the dissociative path.17.21x...1. In our kinetic analysis we have considered four possible rate-limiting steps: H 2 activation ŽEq. 4yx . At high PPh 3 concentrations. xy1 Ž PPh 3 . Furthermore. CH 4 Žas inert balance. Rh concentration Žmmolrl. In fact. PPh 3 concentration Žmmolrl. HRh Ž C 2 H 4 . only the latter is considered.0 562. In accounting for the Rh in solution. CO insertion ŽEq. Ž x s 1 or 2.. x Ž PPh 3 .. C 2 H 5 –Rh Ž CO .2 55. x Ž PPh 3 . we have found that the kinetics of ethene hydroformylation can be described satisfactorily without invoking the associative route.Ž CO . 4yx . the vast majority of Rh is present as HRhŽ CO. where x s 1 or 2.0 37. Ž 13.. HRh Ž CO . ethene insertion ŽEq. The 16-electron initiators of the catalytic cycles are formed by the dissociation of CO or PPh 3 from these 18-electron complexes. and all other species are negligible. 3yx .1 287. 4yx Ž x s 1–3. Ž14. should not be rate-limiting since it has been found to be fast on the time scale of hydroformylation w16. 3yx ™ C 2 H 5 –Rh Ž CO . H 2 partial pressure ŽkPa. in agreement with literature analyses w1. and ethene addition ŽEq. Total reactor pressure ŽkPa. .168 G. thus the rate-limiting step has to be prior to H 2 activation. x Ž PPh 3 . Ž 14. is also low at the reaction conditions of the present study. C 2 H 5 CO–Rh Ž CO . Indeed. .2x.5 1100 93 0. 3yx . that the form of the kinetic expression would not be different for any other neighboring combinations Ž i. We have excluded the two extreme combinations Ž x s 0 and 3. x Ž PPh 3 . however. Kinetic model for ethene hydroformylation As we discussed in Section 3. We have also rejected CO insertion as the rate-limiting step.. As we mentioned earlier. the kinetic order for H 2 is zero.. x s 2 or 3 x s 1 or 2 x s 1 or 2 x s 1 or 2 Ž 11.. Ž12. since their concentrations at our conditions are expected to be much lower than for x s 1 or 2..e. 3yx . and there is no reason to assume that they would significantly contribute to the conversion of ethene. 4yx ™ C 2 H 5 CO–Rh Ž CO . 0 and 1 or 2 and 3. at the conditions of our modeling study. 3yx .r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Table 1 Process variable ranges in the kinetic model database for the hydroformylation of ethene with RhrPPh 3 catalyst in tetraglyme Variable Temperature Ž8C. and Ž 13. all of our kinetic observations can be reconciled with a refined Wilkinson mechanism in which two parallel catalytic cycles are initiated by the coordination of ethene to the 16-electron hydrides with the general formula of HRhŽCO. since it could not be reconciled with a negative kinetic order for CO. partial pressure ŽkPa.932 3.0 7. we can assume that only the coordinatively saturated components are present in significant concentrations. The rate-determining step therefore is either the addition or the insertion of ethene Ž Eqs. NMR results indicate w14. Minimum 80.Ž CO . Ž 12. CO partial pressure ŽkPa. Ž11. C 2 H 4 partial pressure ŽkPa.e.0 8. Ž 14. Ž13. x Ž PPh 3 . x ŽPPh 3 .0 747. . 3yx q C 2 H 4 ™ HRh Ž C 2 H 4 . Kiss et al.. 3yx q H 2 ™ C 2 H 5 –COH q HRh Ž CO . It can be shown.8 853 1252 372 0.1. x Ž PPh 3 .

Mechanism for ethene hydroformylation with RhrPPh 3 catalyst. K 11. 2 . The total Rh in solution therefore can be accounted for by five mononuclear 18-electron complexes. However. x ŽPPh 3 . it does not affect the final form of the kinetic model since it is the same for both scenarios. the concentration of these dimers must be negligible and therefore they are not included in the Rh-balance.. 3yx Ž x s 1 or 2. as will be shown later. Chapter 3. Ž15. This mechanism yields the following rate expression for ethene hydroformylation: k 1 K 1 K 11 TOF s 1 q K 1 K 11 pC 2 H 4 PPh 3 pC 2 H 4 PPh 3 pCO q k 2 K 2 K 21 qK2 K3 pC 2 H 4 pCO PPh 3 pCO pC 2 H 4 pCO Ž 15.. Kiss et al. and K 21 are equilibrium constants and k 1 and k 2 are rate constants.G. leads to zero order in ethene. However. Since the k 1 and k 2 paths are parallel. For further analysis ethene insertion is chosen as a rate-determining step. 4y x Ž x s 1–3. Since all reactions take place in the liquid phase. or HRhŽC 2 H 4 .1. the conditions of the modeling database cover the ethene first order kinetic regime only and the analysis will be limited to this region.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 169 respectively. These complexes are at equilibrium via rapid exchange of C 2 H 4 . the overall rate is the sum of two rates. . This choice will conveniently allow the discussion of both cases and will not influence the final results. and PPh 3. which has been observed experimentally for the high ethenerlow PPh 3 case Ž cf.ŽPPh 3 . K 3 .. x Ž PPh 3 . At these conditions the concentration of Scheme 2. CO. 2 Ž PPh 3 . It should be noticed that at high ethene and low PPh 3 or CO concentrations Eq. Our kinetic data cannot resolve the two reactions but. HRhŽ CO. K 2 . K 1. The Rh-species reacting in the rate-determining step are HRhŽ C 2 H 4 .Ž CO. These Henry’s coefficients are assumed to be independent of the gas phase composition. . ŽCO. The mechanism derived from the previous analysis is depicted in Scheme 2. and HRhŽ C 2 H 4 . for each of the gaseous components. The scheme includes a Rh-dimer to show how H 2 can increase the concentration of the catalytically active mononuclear Rh-species.Ž CO. the equilibrium constants will implicitly contain the vapor–liquid equilibrium proportionality ŽHenry’s coefficient. q K1 K0 PPh 3 q K 2 K 21 where K 0 . since the process-variable range covered in our modeling database falls into the zero order regime for H 2 .

. Ž 16. kX2 .rŽmol Rh s. 3. Without the kX1 term X ŽCase 1. have been regressed using all data Ž 44 data points. The results from three regression cases Ž kX1 s 0. is first order in pC 2 H 4 and wRhx. kX2 / 0. and K 2 s K 2 K 3.2. Thus it is justifiable to assume that kX2 is 0. negative value for kX2 . Eq. monotonously decreasing with wPPh 3 x and has a maximum with respect to pCO . The model therefore correctly describes the kinetic correlations observed experimentally Ž cf. kX2 s k 2 K 2 K 21.6%.409 0. are shown in Table 2.000345 0. in the modeling database at 1008C.401 TOF in Žmol aldehyde. in turn. intermediate.. The final kinetic expression therefore simplifies to: pC 2 H 4 kX1 PPh 3 TOF s Ž 17. or preceeds the formation of a 16-electron alkyl intermediate Ž see Scheme 2. and the contribution of the path through HRhŽC 2 H 4 .r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 the ethene complexes HRhŽ C 2 H 4 . X X The four constants in Eq. gives an excellent fit. and kX1. . kX2 s 0. The kX1 route alone X ŽCase 2. Section 3. That 16-electron Table 2 Regression results for Eq. wPPh 3 x in molrl. Ž 15. kX1. Kiss et al. pCO PPh 3 X 1 q K1 qK2 PPh 3 pCO The experimental data and the fit by the above kinetic model are shown in Fig. K 1. the lack of equilibrium data does not allow the resolution of the individual equilibrium constants.0 31.0176 0.000496 0.8 2. Furthermore. K 1 s K 1 K 0 . . It can be easily shown that the final kinetic expression would be the same as Eq. 3yx Ž x s 1 or 2.53 X ÝŽerror. the standard error is high. Mechanistically. the ethene-containing terms in the denominator are negligible. therefore can be simplified to Eq. and results in a non-physical. brings no improvement in the standard error. where the average error is 7. 2 is negligible. if the rate-determining step were ethene-addition.000473 X K2 396. 2 Ž PPh 3 .Ž CO. Ž16. K 2 . . The inclusion of k 2 ŽCase 3.9 0. Ž 16. Ž16.Ž CO . pCO PPh 3 X X 1 q K1 qK2 PPh 3 pCO X X where kX1 s k 1 K 1 K 11. Pressures in kPa.0173 X k2 7. Kinetic model for hydrogenation selectiÕity in ethene hydroformylation In the kinetic analysis of hydroformylation we have concluded that the rate-determining step either is.170 G. Ž 17. and the equilibrium constant K 1 is negative.1.: pC 2 H 4 pC H kX1 q kX2 2 4 PPh 3 pCO TOF s Ž 16.Ž CO.68 0 y0. using CSTR data for ethene hydroformylation with RhrPPh 3 catalyst at 1008C k1 Case 1 Case 2 Case 3 0 0.ŽPPh 3 .2. this implies that the main pathway for ethene hydroformylation is through the HRh Ž C 2 H 4 .318 X K1 y0. must be low and. x Ž PPh 3 . 13. zero order in p H 2 . Eq. The regression results suggest that the dominating conversion path is kX1. 2 80. as well.

10. the hydrogenation vs. Based on our previous discussions. or H 2 . . that is what was found experimentally Ž see Fig. . Mechanism for ethene hydrogenation under hydroformylation conditions with RhrPPh 3 catalyst. PPh 3 . Scheme 3. Furthermore. on the other hand. Kiss et al. 13.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 171 Fig.1. The former two reactions would yield acyl-intermediates after CO insertion. intermediate can react with CO.G. H 2 for the unsaturated alkyl intermediate. There are two branching points between hydroformylation and hydrogenation. By this mechanism the rate of hydrogenation should be first order in H 2 . hydrogenation selectivity is determined by the competition of CO and PPh 3 vs. and ultimately lead to the formation of the aldehyde Ž Scheme 1. Section 3. since that is the dominating path under the conditions of this study Žcf.. however. Model fit for the hydroformylation of ethene with RhrPPh 3 catalyst in tetraglyme at 1008C. will yield ethane. only the route via k 1 needs to be considered.2. Therefore the hydroformylation vs. . and indeed. Scheme 3 depicts all possible routes leading to propanal and ethane. The addition of H 2 to the coordinatively unsaturated alkyl intermediate.

172 G. 3.3. to determine the temperature dependence of the equilibrium constants. The regressed value for kX at 1008C is 0. there are not enough data available for a full regression of Eq. 14. and Eq. 14. . However. Temperature effects At temperatures other than 1008C. ethane selectivity is found to be a linear function of p H 2rpCO and independent of the concentration of PPh 3 ŽFig.4 kcalrmol for hydrogenation. since the heat of reaction for the COrPPh 3 ligand-exchange is expected to be small. and can be simply expressed as: TOFC 2 H 6 TOFC 2 H 5CHO s kX pH2 pCO Ž 19. Model fit for the hydrogenation vs. Experimentally. hydroformylation selectivity is determined by the competition between H 2 and CO Ž k 1H vs. hydroformylation selectivity of ethene with RhrPPh 3 catalyst at 1008C. provides an excellent fit over more than two orders of magnitude. as shown by Eq.00202. Assuming therefore that K 1 and K 2 are not temperature dependent will likely give an acceptable fit. Applying the above approximation in a regression of all the data in the Appendix yields activation energies of 17. X X Ž17. . These values are indeed not much different from the ones obtained experimentally at constant Fig. K 1 and K 2 probably are not strong functions of X X temperature. and the gas X X solubilities are not strongly temperature dependent. k 12 route.: TOFC 2 H 6 TOFC 2 H 5CHO s p H 2 k 1H pCO k 12 q PPh 3 k 11 Ž 18. K 1 and K 2 .2.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 hydroformylation selectivity is expected to be the function of the relative concentrations of H 2 vs. Ž19.0 kcalrmol for hydroformylation and 20. CO and PPh 3 . Ž18. Kiss et al. It suggests that the hydrogenation vs. .

2 and 19. The aldehyde hydrogenation selectivity is very low. Kiss et al. Ethane selectivity increases with increasing temperature. However. Harry W. indicating that the activation energy for olefin hydrogenation is higher than hydroformylation.. s 0. Under typical hydroformylation conditions the selectivity for propionaldehyde is nearly 100%. when ethene saturation occurs: increasing hydrogen pressure increases the reaction rate by expanding the ethene first order regime. when the PPh 3 concentration is relatively high Ž) 50 mmolrl.00202e 1. The model has proved to be very accurate even at operating conditions which are outside of its database. a kinetic model has been developed to predict the rates of ethene hydroformylation and hydrogenation..0176e T TOFPropanal s 1. Roy L.rŽmol Rh s. is first order in rhodium. CO for a 16-electron alkyl intermediate in a post-rate-determining step.000496 and TOFEthane TOFPropanal 3400 pCO PPh 3 PPh 3 PPh 3 q 31. Our kinetic data suggest that the rate-determining step in the ethene first order regime is olefin addition or olefin insertion. TOF is in Ž mol product. Based on a refined Wilkinson mechanism w1x. pressures are in kPa. This phenomenon may be linked to a shift in the rate determining step of the reaction. Deckman for the many useful discussions..1 kcalrmol for hydroformylation and 22 kcalrmol for hydrogenation. and wPPh 3 x is in molrl. Propanol can only be detected in long term CSTR runs. The kinetic order for ethene is also one at the industrially important conditions.4 mol%. It does have an effect. In this model the hydrogenation vs.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 173 COrPPh 3 ratios Žsee Chapter 3. The olefin hydrogenation selectivity is low Žapproximately 0. Activation energies obtained at constant PPh 3rCO ratios range between 18. below which a positive fractional order can be observed. Hydrogen has little effect on the rate of hydroformylation in the ethene first order kinetic regime. The kinetic effect of the CO and PPh 3 ligands is also a function of their concentration: both inhibit the reaction above a threshold value. Conclusions The kinetics of ethene hydroformylation with RhrPPh 3 catalyst ŽPrRh) 3. We would also like to thank Dr. 4.1. the combination of low phosphine and high olefin concentration leads to a saturation in ethene. and can be described by a linear function of the p H 2rpCO ratio. Pruett and Dr.2–0.987 ž 1 y 373 T /p H2 pCO Ž 21. The hydroformylation rates and hydrogenation selectivities at other than 1008C thus can be calculated as follows: 17000 0.987 ž 1 y 373 1 T / pC 2 H 4 1 q 0.G. Acknowledgements The authors would like to thank the management of Exxon Research and Engineering and Exxon Chemical for their support. hydroformylation selectivity is determined by the competition of H 2 vs. however.8 pCO 1 Ž 20. .

7 486.2 495.0 100.059 0.5 533.0 14.0245 0.9 207.34 0.2 18.05 0.0 29.5594 1.233 0.3982 0.15 0.0006 0.4 218.2 213.61 0.0 80.0 8.0006 0.0195 0.127 PPh 3 mmolrl 94 94 94 94 189 189 190 190 191 191 94 93 185 185 186 186 186 94 185 93 93 93 93 93 93 93 93 kPa abs 1234 1227 1222 1231 1105 1105 1105 1105 1105 1105 1227 1235 1249 1249 1247 1247 1244 1229 1247 1221 1222 1222 1238 1223 1221 1223 1222 Total pressure Partial pressures in the reactor CO kPa 211.147 0. 0.0 90.rŽŽmol Rh.4 499.65 0.0 90.5 54.08 0. s.0016 0.9495 0.0 100.0408 0.0 90.9 545.39 0.6 26.3 209.8 336.8 596.4 459.127 0.117 0.5 580.0 33.9 15.6 506.6 35.924 0.5 539.0054 0.0 90.0 100.4 31.3 444.3 30.0 595.2 27.7 30.0062 0.2 602.0 100.116 0.4 11.8339 0.14 0.8 40.13 2.6 223.9 30.3 585.2 600.0 100.930 0.0 11.5895 0.0 100.0006 0.9 27.7 30.8 338.4893 0.117 0.0 C2 H4 kPa 20.2016 .45 0.2381 1.3 CH 4 kPa 382.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 Appendix A.0 90.0 90.6 103.0024 0.1313 1.3 387.2 25.4 55.0 100.7 33.4 9.0 90.3 207.7 17.07 0.4703 0.0 100.0 576.8 209.1586 0.0053 0.6 18.146 0.5 206.2 381.1 576.3 605.7 15.9252 0.8 489.0578 0.9 12.7 12.4 39.8 18.48 Propanal kPa 21.0392 0.1696 1.1 25.2 551.8 596.0820 0.2 109.7 446.174 G.8864 1.118 0.7 720.0056 0.0 24.6429 0.16 0.0173 0.0182 Propanal Žmol product.0097 0.7 13.0 90.1 30.7 29.0 90.0060 0.932 0.8 9.3 376.3 717.8 454.8 373.04 0.127 0.06 0.4 12.4 34.932 0.37 1.2 29.9 22.0 100.5901 0.0050 0.0 mmolrl 0.0 100.059 0.5430 0.1 596.8 9.3302 1.9544 0.0021 0.6 73.116 0.0 504.9 H2 kPa 598.08 1.3 29.13 1.234 0.67 0.0 30.127 0.127 0.1698 1.7 213.5 522.8 55.9 500.6 48.0 100.5 56.4991 0.8 57.927 0.4 716.6215 0.5 33.4 13.2 594.8 486.3 32.7983 1.0 645.8 24.147 0.0257 0.146 0.18 0.6 9.7 19.0 100.0083 0.7 529.3 551.1 31.0 453.9 506.6 71.146 0.0026 0.06 0.0 100.8 379.5 30.5 381.0042 0.0 90.3 13.926 0.3 38.0111 0.43 3.7 595.4 9.51 0.6162 1.0278 0.7 353.1 23. Kiss et al.6447 0.16 0.0 100.6 598.9 387.7 31. Continuous flow kinetic data for the hydroformylation of ethene with RhrPPh 3 catalyst in tetraglyme r Temperature Liquid concentration Rh 8C 80.0 387.22 0.146 0.0 573.6 208.127 0.0 TOF C2 H6 0.5 19.2023 1.3 C2 H6 kPa 0.1448 1.3 209.11 0.5 12.7 37.14 0.0 100.4 650.

0 100.0 746.2 374.175 0.467 0.0 100.0118 0.116 93 93 93 94 94 94 94 183 185 185 185 186 186 186 188 188 189 190 367 367 368 371 372 372 187 183 185 186 186 93 93 92 1223 1223 1224 1222 1220 1221 1221 1226 1235 1217 1226 1222 1235 1226 1100 1100 1100 1100 1221 1223 1221 1222 1221 1222 1247 1240 1246 1248 1252 1223 1232 1229 412.128 0.0 653.4 36.8 337.127 0.0038 0.9 95.7 206.0046 0.04 0.8 598.5 23.2 40.0740 0.3 101.20 0.4 58.127 0.0 100.0 100.0045 1.2 385.0597 0.19 0.2740 1.3129 G.5 12.0143 0.0 100.7 27.1981 1.1 32.1 34.1062 1.8763 1.0 100.6 12.129 0.0150 0.07 0.27 0.6 616.6 125.9823 0.0234 0.9 30.5 37.127 0.0 28.4 13.0 100.0 13.7 119.100.0065 0.6 48.0039 0.4 0.08 0.115 0.8421 1.8 35.0 100.06 0.0 110.8 209.0 100.9 514.0 612.6 27.9 14.1 343.4 535.3 16.0 100.0024 0.7 565.1983 0.3 578.0038 0.2 18.6 399.9 339.1 12.116 0.2 491.2 31.8 33.20 0.8 591.33 0.2 574.0 100.7 0.7 203.5 542.0014 0.08 0.2 206.0 607.6 490.0231 0.9 26.1535 1.44 1.2 211.9 491.117 0.1 23.1 594.1 371.5 484.0015 0.3 779.2 546.058 0.2 41.10 0.1 31.7 561.9319 0.4806 0.20 0.472 0.08 0.9 36.0052 0.54 0.38 2.9 781.5 499.8 59.2 67.0125 0.8 69.0063 0.144 0.0015 0.0 29.470 0.8 16.8 15.1 30.4 8.176 0.1430 1.0262 0.174 0.13 0.0 100.0233 0.0009 0.2 488.74 3.1042 1.1 714.8 569.0 12.9 11.9 214.1 11.7 367.0 100.1 14.3 169.9787 0.1 356.5 13.8 33.127 0. Kiss et al.177 0.9 13.0 110.6 371.2129 1.4 486.0 100.0 352.0104 0.3 365.1361 0.0 100.9 535.6 482.4 608.8593 0.0 360.14 0.0 100.8 16.8 30.0251 0.0068 0.23 0.177 0.7 608.6 217.0 100.0202 0.7764 2.12 0.9 43.0021 0.0 100.6 8.5 120.6 18.4 23.8 14.7 15.0 13.7 10.3 29.0 100.0 100.5 20.9 28.0 100.9 262.18 0.6 32.2 547.8499 0.0 621.0 100.2776 1.21 0.117 0.7 624.8 391.4 543.147 0.8 147.6589 1.0018 0.6 359.3 536.4 15.0018 0.4 31.0073 0.8 12.129 0.0 100.12 0.9 35.8 30.4 553.0268 1.7 16.0 100.3 464.2 783.177 0.2 852.6 42.38 0.4 437.6 16.146 0.23 28.8008 1.4 46.9 30.0 100.8574 0.8 539.3 17.08 0.9 204.60 0.18 0.0 110.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 175 .0 44.176 0.04 0.9 127.0 100.468 0.4267 0.3 426.9 286.6756 0.128 0.9 131.2 267.6243 0.129 0.147 0.1 216.42 1.8272 1.4 40.7 484.2 386.6 11.4 43.0 100.0098 0.0073 0.0 100.0222 0.129 0.8 22.1 0.09 0.0086 1.147 0.0 565.2 142.6852 0.3 552.8 28.

Meeting. Matsumoto. Res. J. Oswald. J.M. Sidebottom. 35 Ž1996. Ind. Chaudhari.G. Brown. Org. 327. Robinson. 1996. J. CHEMTECH Ž1972. J.P. w23x M. Mozeleski. 5 Ž1979. Kent. Iggo. Kobyakov.A.K. 431. R. G. Mozeleski. 3133.M. G. R. 292.A. Papile. Inorg.. Unruh.W. Catal. Ind. Soc. S. in: B. Catal. J.S.M. T. Soc.V. M.J.J. A. Petr. Chaudhari. Chem. 2 Ž1989. 4438. J. 1983. 33 Ž5. ŽA. Catal. 78 Ž1993.R. w11x D.A. Breslow. J. Chem. Duwell. 2753. Deshpande. 115 Ž1989. J. Commun. w16x R. K. M. Korneyeva.A. H. Chem. Brannon.. W. w32x V. Sakulin..E. D. 721. Mol. Herrmann ŽEds. Ž1969. J.D.T.M. Hydrocarbon Processing Ž1970.H. Congr. J.E. w19x R.V. Soc.S. Ye. Reisch. Lett. York. Chem. w9x R. Preprints Div.A. 572. Chaudhari. H. Chem. w40x A. 209. Kastrup.S. ŽLondon. Sanger. Divekar. Res. F. Reisch.M. 1996. 1392.V. Purwanto. J. 3927. Kastrup. 554.M. J.M. Canning. G. ACS Natl. Pruett. 85.B. J. Ser. 265. 247. in: L. Heck.V. C. R. Loktev. C.. Ž1968. 28 Ž1989.. w37x M.M. R. Ekstrom. w27x B. Soc.R. 125. Miller. J. Osborn.L.. G. Ye. Ž1993. Loktev. w5x G. Raimondi. Kastrup. Montrasi. Chem. ŽMilan. ŽA. w15x A. Mol. J. Mol. 41 Ž1987. NV. VCH. Wilkinson. Evans. King. Heck. 4023. A. Evans. 3 Ž1978.J.C. w7x R.V. J. 34 Ž5. Delmas. 82 Ž1960. G. Perkin Trans.A. Hjortkjaer. Soc. E. J. E. 271. Chaudhari. 19. Brown. 191.G. D. 1725. 171 Ž1981. Chaudhari. R. Ž1970. Chem. Res. Wilkinson.E. J. D. G. Chem. R. Cornils.D. Chem. Merola. 393. Catal. M. Smith. Strohmeier.. 372.L. Homogeneous Catalysis with Metal Phosphine Complexes. R. Heck. Andreetta. Ž1968. Chem. R.S. Mol. Tsang.S.M.. w8x D. Whyman. 377.F. Oswald. Lett. S. R. Am. Garland. R. Catal. 8 Ž1969. ¨ w42x J. Indian Chem. w33x R. A. Aronovich. Claridge. Deshpande. 196 Ž1982. Oswald. Khromushina. Chem. Petr. Deshpande. 26 Ž1987. Weininger. G Wilkinson.r Journal of Molecular Catalysis A: Chemical 138 (1999) 155–176 References w1x C. Wilkinson. Chaudhari. E.176 G. J. J. G. w26x P. Catal. 326.V. w48x A. w18x J. Kohlpaintner. G.H. Horvath. Chem. 89 Ž1994. B. Soc.S. 27 Ž1988. J. Douthwaite. Brown.S. w6x C. J.F. Heaton. J.V. B. New York. Oswald.C. Proc. J. R. Chem. 113. w12x T. G. w46x R. w41x W. w47x K. Tinker.M. Gholap.Y. Tetrahedron Lett. Catal. Res. Chem. Kurkin.F. Christenson. 291. Ueda. I&EC Prod. Spring 27Ž2. Eng. Olivier. L37. w35x R. Basner. Chim. Bhanage. Ind. Kent. J.F.A. 8 Ž1969. R. A. J.A.I.K. 12 Ž1981. R. Faller. Res. 1168. 1. Gregorio. w4x D. w22x I. Pignolet ŽEd.V.D.A. S. Tolman. 2660. King. w39x J. Eguchi. Divekar.A. Vladimirova. 124.A. w21x J. Petr. D.T. Hughes. Mol. Hydroformylation Žoxo synthesis. Slivinskii. Mechanistic studies of catalytic reactions using spectroscopic and kinetic techniques.M.R. Hendriksen. Brown. Potarin. A. w29x S. A. 5th Int. Eng.V.R. Merola. 2499. Bhattacharya. Res. Divekar.S. Green. J. . Ž1991. Breslow.S. Mol. Booth. Catal. 1 Ž1973. Chem. Chem. Ind. Brown. 503. w30x R. Soc.B.A. Ž1982. Breslow. Mol. Kiss et al.B. A. A.H. 83 Ž1961. Eng. Chem. K. A.H. Chem. 112. Applied Homogeneous Catalysis with Organometallic Compounds. ŽA.M. M. Catal. w34x J. 62 Ž1982. Catal.V. 115 Ž1997.V. Yagupsky. Me. V.V.F. Pagani. Adv. Craddock. L. Am. w20x W. Chem. Hershman.R. Chem. Slivinskii. Vol. Bhanage. II Ž1987. Catal. Brown.T.M. w45x A. Eng. T. 101 Ž1979. Roth. J. S. Mol. Craddock. Ž1982. ACS Symp. Ž1960. w31x R. Chaudhari. 467.E. Las Vegas.V. D. J. Frohning. King. Foster. Chem. S. 391. J. 21 Ž1993. 84 Ž1962.A.A. Wilkinson. R. w44x S.V.J.J. w3x D. Chem. Tamura. S. Soc. w24x D. 410. J. Divekar. L. 723. 71. Al. Dev.Z.V. Chem. Iqbal. w13x R. Dev. Data 35 Ž1990. 34 Ž1969. Catal.S. w36x J.M. J. Chem. d’Oro.L.M. J. Ch. w43x C. w25x G. Chem. Soc. 16 Ž1982. Ser. J. J..C. J. J. Tanaka.A. Yagupsky. R. Heck.V. Soc. N. 1389. Breslow. 14 Ž1982. Ž1970. 4893. Deshpande. Hershman. 43. A. Ind. Plenum. 1597. w17x J. Ž1982. Am. J.I. P. New York. w38x O. Catal. Kent. R. Soc. w10x R. Roth.R.F. Chaudhari. Lago. Am. 221. Bor. R. Morris. Deshpande. ŽA. 30 Ž1991. 110 Ž1976. I&EC Prod. Chem. R.M. A. 677. Chem.B.G. Organomet. Paulik.D. F. G. Deshpande.K. Mol. S. Ind. w2x C. 59 Ž1982. W. A: Chem. R. Dalton Trans. Kuhn. Commun. w14x A. Deshpande. J.K. Kastrup.V. Brown. Korneyeva.C. w28x R. Moser.K. Merola. Roelen reaction. Soc.. R.E. Unruh.B. Soc. J. Ž1994.S. Eng.